diff --git a/README.md b/README.md
index c95dfa8..c282cc2 100644
--- a/README.md
+++ b/README.md
@@ -1,450 +1,244 @@
 # BrowserStack for Cursor
 
-Test websites and mobile apps on 3,500+ real devices and browsers—directly from Cursor using natural language. No context switching and very easy to setup.
+Access BrowserStack's complete testing platform directly from Cursor using natural language. Test on 3,500+ real devices and browsers, run automated tests, scan for accessibility issues, and generate test cases from requirements—all without leaving your IDE.
 
-## What You Can Do
+## Getting Started
 
-**Test anywhere, from anywhere**
-- Open your localhost app on an iPhone 15 Pro Max
-- Test your website on Safari 17 without owning a Mac
-- Debug app crashes on Android 14 in real-time
-- Run Playwright tests on 50+ browser/OS combinations
+### Prerequisites
 
-**Stay in your flow**
-- Ask in plain English: *"Test my site on Edge"*
-- Get instant access to real devices
-- Debug failed tests without leaving Cursor
-- Fix accessibility issues with AI suggestions
+- Node.js 18 or higher ([download](https://nodejs.org/en/download))
+- BrowserStack account ([sign up for free trial](https://www.browserstack.com/users/sign_up))
 
-**Automate everything**
-- Run test suites on BrowserStack infrastructure
-- Generate test cases from product requirements
-- Auto-heal flaky tests with AI
-- Manage test cases and track results
+### Installation
 
-## Setup
+1. **Get your BrowserStack credentials** from [Account Settings](https://www.browserstack.com/accounts/profile/details)
 
-### Step 1: Install Node.js
+2. **Configure MCP settings in Cursor**:
+   - Open Cursor Settings
+   - Navigate to MCP configuration
+   - Add BrowserStack credentials:
 
-You need Node.js version 18 or higher (we recommend v22.15.0 LTS).
-
-Check your version:
-```bash
-node --version
-```
-
-If you need to install or update:
-- **macOS**: `brew update && brew upgrade node` or [download here](https://nodejs.org/en/download)
-- **Windows**: Download from [nodejs.org](https://nodejs.org/en/download)
-
-### Step 2: Get BrowserStack Credentials
-
-1. **Sign up** at [browserstack.com/users/sign_up](https://www.browserstack.com/users/sign_up)
-   - Free trial available, no credit card required
-   - Open source projects get free access
-
-2. **Get your credentials** from [Account Settings](https://www.browserstack.com/accounts/profile/details)
-   - You'll need your `Username` and `Access Key`
-   - Keep these handy for the next step
-
-### Step 3: Install the Plugin
-
-1. **Install from Cursor Marketplace** (or manually)
-
-2. **Add your credentials** to your environment:
-
-   **On macOS/Linux:**
-   ```bash
-   # Open your shell profile
-   nano ~/.zshrc  # or ~/.bashrc if you use bash
-   
-   # Add these lines at the end (replace with your actual credentials)
-   export BROWSERSTACK_USERNAME="your_username_here"
-   export BROWSERSTACK_ACCESS_KEY="your_access_key_here"
-   
-   # Save and reload
-   source ~/.zshrc
-   ```
-
-   **On Windows (PowerShell):**
-   ```powershell
-   # Open PowerShell profile
-   notepad $PROFILE
-   
-   # Add these lines
-   $env:BROWSERSTACK_USERNAME="your_username_here"
-   $env:BROWSERSTACK_ACCESS_KEY="your_access_key_here"
+   ```json
+   {
+     "mcpServers": {
+       "browserstack": {
+         "command": "npx",
+         "args": ["-y", "@browserstack/mcp-server@latest"],
+         "env": {
+           "BROWSERSTACK_USERNAME": "${BROWSERSTACK_USERNAME}",
+           "BROWSERSTACK_ACCESS_KEY": "${BROWSERSTACK_ACCESS_KEY}"
+         }
+       }
+     }
+   }
    ```
 
-3. **Restart Cursor** to activate the plugin
-
-4. **Verify installation** by asking Cursor:
+3. **Restart Cursor** and verify:
    ```
    "Open google.com on Chrome"
    ```
 
-✅ That's it! You're ready to test.
-
-## Common Workflows
-
-### 🚀 Quick Manual Testing
-
-Testing your local development site on different browsers:
-
-```
-💬 "Open my website running on localhost:3000 on Safari 17"
-💬 "Test localhost:8080 on the latest Chrome on Windows 11"
-💬 "Open my site on Firefox on macOS"
-💬 "Take a screenshot of this page"
-```
+## What's Included
 
-**Use case**: You're developing a feature and need to check if it works on Safari, but you're on Windows.
+### Skills
 
-**Pro tip**: If you find a visual bug, use the **Visual Issue Fix** skill to get browser-specific fixes.
+- **scan-and-fix-accessibility** - Scan webpages for WCAG violations and get code fixes
+- **run-web-tests-on-browserstack** - Run Selenium/Playwright/Cypress tests on multiple browsers
+- **run-mobile-tests-on-browserstack** - Run Appium/XCUITest/Espresso tests on real devices
 
----
+### Agents
 
-### 📱 Mobile App Testing
+- **test-case-generator** - Generate test cases from PRD documents using AI
 
-Test your mobile app on real devices:
+### MCP Tools (20 Available)
 
-```
-💬 "Open my app on iPhone 15 Pro Max"
-   (Cursor will ask for your .ipa or .apk file path)
+**Test Management**
+- `createProjectOrFolder` - Create Test Management projects and folders
+- `createTestCase` - Add manual test cases to projects
+- `listTestCases` - List test cases with filters (priority, status, tags)
+- `createTestRun` - Create test runs for selected test cases
+- `listTestRuns` - List test runs with date/assignee filters
+- `updateTestRun` - Update test run status, tags, notes
+- `addTestResult` - Add execution results (passed/failed/blocked/skipped)
+- `createTestCasesFromFile` - Bulk create test cases from uploaded files
 
-💬 "Test my app on Galaxy S24 with Android 14"
-💬 "My app crashes on iOS 17 when I tap the login button, help me debug"
-```
+**Automate (Web Testing)**
+- `setupBrowserStackAutomateTests` - Integrate SDK and run web tests
+- `fetchAutomationScreenshots` - Get screenshots from test sessions
+- `getFailureLogs` - Retrieve error logs for failed tests
 
-**Use case**: You don't have physical access to an iPhone 15 Pro Max but need to verify the login flow works.
+**App Automate (Mobile Testing)**
+- `takeAppScreenshot` - Launch app and capture screenshot
+- `runAppTestsOnBrowserStack` - Run automated mobile tests on real devices
 
----
+**Live Testing**
+- `runBrowserLiveSession` - Start manual browser testing session
+- `runAppLiveSession` - Start manual app testing session on real devices
 
-### 🧪 Running Automated Tests
+**Accessibility**
+- `startAccessibilityScan` - Scan webpages for WCAG violations
+- `accessibilityExpert` - Get WCAG guidelines and best practices
 
-Run your existing test suite on BrowserStack:
+**AI Agents**
+- `uploadProductRequirementFile` - Upload PRD/PDF for test case generation
+- `fetchSelfHealedSelectors` - Get AI-healed selectors for flaky tests
+- `createLCASteps` - Convert manual test cases to low-code automation
 
-```
-💬 "Setup my Playwright tests to run on BrowserStack"
-💬 "Run my tests on Chrome, Firefox, and Safari"
-💬 "My test suite failed on BrowserStack session abc123, help me debug"
-```
+## Use Cases
 
-**Workflow example**:
-1. You: *"Setup my Playwright tests to run on BrowserStack"*
-2. Cursor updates your config files automatically
-3. You: *"Run my tests on Chrome and Edge"*
-4. BrowserStack runs your tests, Cursor shows you results
-5. If tests fail: *"Help me debug the failures"*
+### 1. Generate Test Cases from Requirements
 
----
+**What**: Upload PRD documents and automatically generate comprehensive test cases with AI.
 
-### ♿ Accessibility Scanning
+**Agent**: `test-case-generator` (invoke with `@test-case-generator`)
 
-Catch accessibility issues before production (uses **Accessibility Fix** skill):
+**Tools**: `uploadProductRequirementFile`, `createTestCasesFromFile`
 
+**Examples**:
 ```
-💬 "Scan accessibility issues on localhost:3000"
-💬 "Fix accessibility problems on my checkout page"
-💬 "Check WCAG 2.1 AA compliance for my signup form"
-```
-
-**Use case**: You want to ensure your signup form is accessible before deploying.
-
-**What happens**: The skill scans your page, identifies violations, provides specific code fixes, and verifies after you implement them.
-
----
-
-### 📋 Test Case Management
-
-Organize and track your testing:
+@test-case-generator
+[Attach your PRD file]
+"Generate test cases for Test Management project PR-12345"
 
-```
-💬 "Create a Test Management project called 'Payment Flow'"
-💬 "Add a test case: Verify credit card payment with invalid CVV"
-💬 "List all high priority test cases in Payment Flow"
-💬 "Create a test run for smoke tests"
-💬 "Mark test case TC-123 as passed"
+@test-case-generator
+"Upload PRD from /Users/xyz/login-flow.pdf and generate test cases for project PR-54321"
 ```
 
-**Workflow example**:
-1. Create project structure
-2. Add test cases as you develop features
-3. Create test runs before releases
-4. Track results and generate reports
+**What you get**:
+- Test cases with title, steps, expected results, priority, and tags
+- Automatically added to your Test Management project (PR-xxxxx format)
+- Coverage of happy paths, error scenarios, edge cases, security
 
 ---
 
-### 🤖 AI-Powered Testing
+### 2. Manage Test Cases and Runs
 
-**Generate test cases from requirements** (uses **Test Case Generator** agent):
-```
-💬 @test-case-generator
-   [Attach your PRD file]
-   "Generate test cases for project PR-12345"
-   
-💬 @test-case-generator
-   "Create test cases for checkout flow with credit card and PayPal"
-   "Add to project PR-67890"
-```
+**What**: Organize test cases, create test runs, and track results.
 
-**Auto-heal flaky tests:**
-```
-💬 "My login test keeps failing because the button selector changed. Fix it with self-healing."
-💬 "Get self-healed selectors for session session_abc123"
-```
+**Tools**: `createProjectOrFolder`, `createTestCase`, `listTestCases`, `createTestRun`, `addTestResult`
 
-**Convert manual tests to automation:**
+**Examples**:
 ```
-💬 "Convert my manual test case TC-45 into a Playwright script"
+"Create a Test Management project called 'Payment Flow'"
+"Add a test case: Verify credit card payment with invalid CVV"
+"List all high priority test cases in Payment Flow"
+"Create a test run for smoke tests"
+"Mark test case TC-123 as passed"
 ```
 
 ---
 
-## Built-in Skills & Agents
-
-The plugin includes specialized workflows and AI agents to streamline common testing tasks.
+### 3. Manual Testing on Real Devices
 
-### 🔧 Skills (Multi-Step Workflows)
+**What**: Test your website or app on real browsers and devices on BrowserStack
 
-#### Accessibility Fix
-**What it does**: Scans your page for WCAG violations, identifies issues, and provides code fixes.
+**Tools**: `runBrowserLiveSession`, `runAppLiveSession`
 
-**When to use**: Before deployment, fixing accessibility bugs, ensuring compliance.
-
-**Sample commands**:
+**Examples**:
 ```
-💬 "Scan accessibility issues on localhost:3000"
-💬 "Fix the accessibility problems on my checkout page"
-💬 "Check WCAG 2.1 AA compliance for localhost:8080/signup"
-```
-
-**Workflow**:
-1. Runs accessibility scan on your page
-2. Categorizes issues (Critical/High/Medium)
-3. Provides specific code fixes for each issue
-4. Re-scans after you implement fixes to verify
-
-**Example fix output**:
-```html
-<!-- Missing form label fix -->
-<!-- Before -->
-<input type="email" placeholder="Email">
-
-<!-- After -->
-<label for="email">Email Address</label>
-<input type="email" id="email" aria-required="true">
+"Open localhost:3000 on Safari 17"
+"Test my website on Chrome 120 Windows 11"
+"Open my app on iPhone 15 Pro Max. App: /path/to/app.ipa"
+"Take a screenshot of the current page"
 ```
 
 ---
 
-#### Visual Issue Fix
-**What it does**: Captures screenshots across browsers, identifies visual bugs, and provides CSS fixes.
+### 4. Run Automated Web Tests
 
-**When to use**: Cross-browser layout issues, responsive design bugs, visual regressions.
+**What**: Execute Selenium, Playwright, or Cypress tests across multiple browsers on BrowserStack infrastructure.
 
-**Sample commands**:
-```
-💬 "My navigation is broken in Safari, help me fix it"
-💬 "Compare how my homepage looks on Chrome vs Firefox"
-💬 "Fix the layout issue on Safari for localhost:3000"
-```
+**Skill**: `run-web-tests-on-browserstack`
+
+**Tools**: `setupBrowserStackAutomateTests`, `fetchAutomationScreenshots`, `getFailureLogs`
 
-**Workflow**:
-1. Opens page on target browsers
-2. Captures screenshots for comparison
-3. Identifies visual differences
-4. Provides browser-specific CSS fixes
-5. Validates fix doesn't break other browsers
-
-**Example fix output**:
-```css
-/* Safari flexbox gap issue fix */
-/* Before */
-.container {
-  display: flex;
-  gap: 20px;
-}
-
-/* After - Works everywhere */
-.container {
-  display: flex;
-}
-.container > * {
-  margin-right: 20px;
-}
+**Examples**:
+```
+"Run my Playwright tests on Chrome and Firefox using BrowserStack"
+"Setup my Selenium tests for BrowserStack and run on Safari 17"
+"Run my Cypress tests on Chrome 120 and Edge. Enable Percy."
+"Get screenshots from Automate session abc123xyz"
+"Get error logs from session ID 21a864032a7459f1e7634222249b316759d6827f"
 ```
 
 ---
 
-### 🤖 AI Agent
-
-#### Test Case Generator
-**What it does**: Generates comprehensive test cases from requirements, user stories, or PRD documents using BrowserStack AI.
+### 5. Run Automated Mobile App Tests
 
-**When to use**: Planning new features, creating test suites, converting requirements to test cases.
-
-**Sample commands**:
-```
-💬 @test-case-generator
-   [Add your PRD file or requirements document to context]
-   "Generate test cases for Test Management project PR-12345"
+**What**: Execute Appium, XCUITest, or Espresso tests on real iOS and Android devices.
 
-💬 @test-case-generator
-   "Create test cases for a login feature with email and Google OAuth"
-   "Add them to project PR-54321"
-```
+**Skill**: `run-mobile-tests-on-browserstack`
 
-**Workflow**:
-1. Invoke agent with `@test-case-generator`
-2. Add requirements file to context (or describe feature)
-3. Provide your Test Management project ID (format: `PR-12345`)
-4. Agent generates structured test cases
-5. Test cases are added to your BrowserStack Test Management project
+**Tools**: `takeAppScreenshot`, `runAppTestsOnBrowserStack`, `getFailureLogs`
 
-**What you get**:
-- Test case title and description
-- Priority level (Critical/High/Medium/Low)
-- Preconditions
-- Step-by-step test steps
-- Expected results
-- Appropriate tags (@smoke, @regression, @p0, etc.)
-
-**Example interaction**:
+**Examples**:
 ```
-You: @test-case-generator
-     [Attach: login-feature-spec.pdf]
-     "Generate test cases for project PR-12345"
-
-Agent: I'll generate comprehensive test cases for your login feature.
-
-       Test cases generated:
-       ✓ TC-001: Login - Valid Credentials - Success (@smoke, @p0)
-       ✓ TC-002: Login - Invalid Password - Error (@regression, @p1)
-       ✓ TC-003: Login - Google OAuth - Success (@smoke, @p0)
-       ✓ TC-004: Login - Empty Fields - Validation (@regression, @p2)
-       
-       Added 4 test cases to Test Management project PR-12345
+"Take a screenshot of my app on iPhone 15 Pro. App: /apps/myapp.ipa"
+"Run Espresso tests on Galaxy S21 and Pixel 6. App: /apps/app.apk, Tests: /tests/suite.zip"
+"Run XCUITest tests on iPhone 14 and 15 Pro with iOS 17. Tests: /tests/login.zip"
+"Get error logs from App Automate session abc123, Build ID xyz789"
 ```
 
 ---
 
-## Real Examples
+### 6. Auto-Heal Flaky Tests
 
-### Example 1: Cross-Browser Bug Fix
+**What**: Fix test failures caused by changed selectors using AI self-healing.
 
-**Scenario**: Users report a broken layout on Safari.
+**Tools**: `fetchSelfHealedSelectors`
 
+**Examples**:
 ```
-💬 "Open localhost:3000/dashboard on Safari 17 on macOS Sonoma"
-   → BrowserStack opens a real Safari session
-   → You see the layout is indeed broken
-   
-💬 "Take a screenshot of this page"
-   → Screenshot saved to your project
-   
-💬 "Now open the same page on Chrome to compare"
-   → Opens Chrome session
-   → Layout works fine
-   
-💬 "What CSS properties might cause this Safari-specific issue?"
-   → Cursor analyzes your code and suggests the problem
+"My login test keeps failing because the button selector changed. Fix it with self-healing."
+"Get self-healed selectors for session session_abc123"
 ```
 
-### Example 2: Mobile App Crash Debug
+---
 
-**Scenario**: Your iOS app crashes on launch for some users.
+### 7. Scan and Fix Accessibility Issues
 
-```
-💬 "Open my app on iPhone 14 with iOS 16"
-   (Provide path: /Users/me/builds/app-v2.1.ipa)
-   → App launches successfully
-   
-💬 "Try on iPhone 15 with iOS 17"
-   → App crashes!
-   
-💬 "Get the crash logs for this session"
-   → Cursor retrieves logs
-   
-💬 "Analyze these crash logs and suggest a fix"
-   → Cursor identifies the issue and suggests code changes
-```
+**What**: Identify WCAG violations on your webpage and get specific code fixes.
 
-### Example 3: Automated Test Failure Investigation
+**Skill**: `scan-and-fix-accessibility`
 
-**Scenario**: Your CI pipeline shows test failures on BrowserStack.
+**Tools**: `startAccessibilityScan`, `accessibilityExpert`
 
+**Examples**:
 ```
-💬 "My Automate build 'feature-branch-123' has 3 failed tests. What went wrong?"
-   → Cursor fetches error logs and screenshots
-   
-💬 "Show me screenshots from the failed tests"
-   → Screenshots displayed
-   
-💬 "The login button selector seems wrong. Update my test file with the correct selector."
-   → Cursor updates your test file
-   
-💬 "Re-run the tests"
-   → Tests pass ✅
+"Run accessibility scan for 'localhost:3000'"
+"Scan accessibility issues on www.mysite.com/checkout"
+"What WCAG guidelines apply to form field error messages?"
+"Re-scan localhost:3000 to verify fixes"
 ```
 
-## What's Available
-
-**Manual Testing**
-- Test websites on any browser/OS combination
-- Test mobile apps on 3,500+ real devices
-- Access localhost from cloud devices (no deployment needed)
-
-**Automated Testing**  
-- Run Playwright, Selenium, Cypress, and more
-- Parallel test execution across browsers
-- Get screenshots, videos, and logs automatically
-
-**Test Management**
-- Create and organize test cases
-- Track test execution and results
-- Generate test reports
-
-**Accessibility Testing**
-- WCAG 2.0/2.1/2.2 compliance scanning
-- AI-powered fix suggestions
-- Local and production site scanning
-
-**AI-Powered Features**
-- Generate test cases from requirements documents
-- Auto-heal flaky test selectors
-- Convert manual tests to automation
-- Intelligent failure analysis and debugging
+**What you get**:
+- Detailed violation report (Critical/High/Medium severity)
+- Specific code fixes for each issue (missing labels, color contrast, keyboard navigation)
+- WCAG compliance verification after fixes
 
 ## Troubleshooting
 
-**Plugin not responding?**
-- Verify your credentials are set correctly: `echo $BROWSERSTACK_USERNAME`
-- Restart Cursor after adding credentials
-- Check Node.js version: `node --version` (need 18+)
-
-**"Authentication failed" error?**
-- Double-check your username and access key from [Account Settings](https://www.browserstack.com/accounts/profile/details)
-- Make sure there are no extra spaces in your environment variables
-
-**Can't access localhost?**
-- This works automatically! Just use `localhost:PORT` in your prompts
-- BrowserStack creates a secure tunnel to your local machine
+**Authentication Issues**
+- Verify credentials in [Account Settings](https://www.browserstack.com/accounts/profile/details)
+- Check MCP configuration in Cursor settings
+- Ensure no extra spaces in username/access key
 
-**Behind a corporate firewall?**
-- Contact your IT team to allowlist BrowserStack domains
-- Or ask us about enterprise firewall configurations
+**Plugin Not Responding**
+- Restart Cursor after configuration changes
+- Check Node.js version: `node --version` (need 18+)
+- View MCP server logs in Cursor
 
-## Get Help
+**Need Help?**
+- [GitHub Issues](https://github.com/browserstack/mcp-server/issues) - Report bugs or issues
+- [BrowserStack Support](https://www.browserstack.com/contact) - Platform questions
+- [Documentation](https://www.browserstack.com/docs) - Detailed guides
 
-- **Issues with the plugin**: [Open a GitHub issue](https://github.com/browserstack/mcp-server/issues)
-- **BrowserStack platform questions**: [Contact Support](https://www.browserstack.com/contact)
-- **Documentation**: [BrowserStack Docs](https://www.browserstack.com/docs)
+---
 
-## Pricing
+## How to Sign Up for BrowserStack?
 
-- **Free Trial**: Test drive with limited minutes
-- **Open Source**: Free for qualifying projects ([apply here](https://www.browserstack.com/open-source))
-- **Paid Plans**: From $29/month for individuals, custom enterprise pricing
+1. **Create a free account** at [browserstack.com/users/sign_up](https://www.browserstack.com/users/sign_up)
+   - No credit card required
+   - Instant access to start testing
 
-See [pricing details](https://www.browserstack.com/pricing)
+2. **Get your credentials** from [Account Settings](https://www.browserstack.com/accounts/profile/details) to configure the plugin
diff --git a/agents/test-case-generator.md b/agents/test-case-generator.md
index 2221ba0..1a1ccc9 100644
--- a/agents/test-case-generator.md
+++ b/agents/test-case-generator.md
@@ -1,12 +1,17 @@
 ---
 name: test-case-generator
-description: Generate comprehensive test cases from product requirements, user stories, or PRD documents using BrowserStack AI.
+description: Generate comprehensive test cases from product requirements using uploadProductRequirementFile and createTestCasesFromFile MCP tools. Use when converting requirements to test cases.
 ---
 
 # Test Case Generator
 
 You help users generate test cases from requirements using BrowserStack's AI-powered test case generation.
 
+## MCP Tools Used
+
+- `uploadProductRequirementFile` (Tool #20) - Upload PRD/PDF/screenshot files
+- `createTestCasesFromFile` (Tool #8) - Generate test cases from uploaded files
+
 ## When to Use
 
 - User provides product requirements or user stories
@@ -15,65 +20,66 @@ You help users generate test cases from requirements using BrowserStack's AI-pow
 
 ## Process
 
-### 1. Get Requirements
-
-Ask user to provide:
-- Text description of the feature
-- Or path to PRD file (PDF, image)
+### 1. Upload Requirements File
 
-### 2. Generate Test Cases
+Use `uploadProductRequirementFile` to upload the PRD:
 
-Use BrowserStack AI to generate test cases:
 ```
-"Generate test cases for [feature description]"
-"Generate test cases from PRD file at [file path]"
+"Upload PRD from /Users/xyz/Desktop/login-flow.pdf and use BrowserStack AI to generate test cases"
 ```
 
-### 3. Organize in Test Management
+**What the tool does:**
+- Uploads PDF, screenshot, or requirements document
+- Returns a file mapping ID for use in test case generation
+- Prepares file for AI processing
+
+### 2. Generate Test Cases
+
+Use `createTestCasesFromFile` to generate test cases from the uploaded file:
 
-Create project structure:
 ```
-"Create Test Management project called '[Project Name]'"
-"Add folder '[Feature Name]' to project"
+"Upload test cases from '/Users/xyz/testcases.pdf' to the 'Shopping App' project in Test Management"
+"Generate test cases for project PR-12345 from the uploaded requirements"
 ```
 
-### 4. Review and Add Test Cases
+**What the tool does:**
+- Analyzes uploaded requirements using AI
+- Generates comprehensive test cases
+- Automatically adds them to specified Test Management project
+- Includes title, steps, expected results, priority, and tags
+
+### 3. Specify Test Management Project
 
-The AI will generate test cases with:
-- Test case title
-- Steps to execute
-- Expected results
-- Priority level
+Provide the project identifier (format: `PR-xxxxx`):
 
-Review and add to Test Management:
 ```
-"Add these test cases to Test Management project"
+"Add generated test cases to Test Management project PR-12345"
 ```
 
 ## Example Workflow
 
-**User**: "Generate test cases for a login feature with email and Google OAuth"
+**User**: "@test-case-generator Generate test cases for project PR-12345"
+[User attaches: login-feature-spec.pdf]
 
 **Your approach**:
+1. Upload file: Use `uploadProductRequirementFile` with the attached PDF
+2. Generate: Use `createTestCasesFromFile` to create test cases for project PR-12345
+3. Confirm: Report which test cases were generated and added
+
+**Response**:
 ```
-I'll generate comprehensive test cases for your login feature.
-
-First, let me understand the requirements:
-- Email/password login
-- Google OAuth login
-- Forgot password flow
-- Any specific error scenarios?
-
-Then I'll use BrowserStack AI to generate test cases covering:
-- Happy path (successful login)
-- Error cases (wrong password, unregistered email)
-- OAuth flow
-- Edge cases (empty fields, SQL injection attempts)
-
-Once generated, I'll organize them in Test Management:
-"Create Test Management project 'User Authentication'"
-"Add test cases for email login scenarios"
-"Add test cases for OAuth login scenarios"
+✓ PRD uploaded and analyzed
+✓ Generated 6 test cases for login feature:
+
+Test cases added to project PR-12345:
+- TC-001: Login - Valid Credentials - Success (@smoke, @p0)
+- TC-002: Login - Invalid Password - Error (@regression, @p1)
+- TC-003: Login - Unregistered Email - Error (@regression, @p1)
+- TC-004: Login - Google OAuth - Success (@smoke, @p0)
+- TC-005: Login - Empty Fields - Validation (@regression, @p2)
+- TC-006: Login - SQL Injection Prevention (@security, @p1)
+
+All test cases added to Test Management project PR-12345
 ```
 
 ## Test Case Structure
@@ -84,12 +90,18 @@ Each generated test case includes:
 - **Preconditions**: What needs to be set up
 - **Steps**: Numbered actions to perform
 - **Expected Results**: What should happen
-- **Tags**: @smoke, @regression, @p0, etc.
+- **Tags**: @smoke, @regression, @p0, @security, etc.
+
+## Required Information
+
+To generate test cases, you need:
+1. **Requirements file**: PRD, PDF, or screenshot (uploaded via `uploadProductRequirementFile`)
+2. **Project ID**: Test Management project identifier (format: `PR-xxxxx`)
 
 ## Tips for Better Test Cases
 
-- Be specific about requirements
-- Include edge cases and error scenarios
+- Ensure requirements file is detailed and clear
+- Include edge cases and error scenarios in requirements
 - Mention any integrations (third-party services)
 - Specify target platforms (web, mobile, specific browsers)
 - Note any compliance requirements (security, accessibility)
diff --git a/skills/accessibility-fix/SKILL.md b/skills/accessibility-fix/SKILL.md
deleted file mode 100644
index cddcbdf..0000000
--- a/skills/accessibility-fix/SKILL.md
+++ /dev/null
@@ -1,104 +0,0 @@
----
-name: accessibility-fix
-description: Scan a webpage for accessibility issues, identify violations, generate code fixes, and verify the fixes work. Use when fixing WCAG compliance or accessibility bugs.
----
-
-# Accessibility Fix Workflow
-
-## When to Use
-
-- User wants to fix accessibility issues on a page
-- Need to ensure WCAG compliance
-- Fixing specific accessibility bugs
-
-## Steps
-
-### 1. Run Accessibility Scan
-
-Ask for the URL to scan:
-```
-"Run accessibility scan on [URL]"
-```
-
-Example: `"Run accessibility scan on localhost:3000"`
-
-### 2. Analyze Results
-
-Review the scan results and categorize issues:
-- **Critical**: Blocks users (missing form labels, keyboard traps)
-- **High**: Major barriers (poor color contrast, missing alt text)
-- **Medium**: Best practices (heading hierarchy)
-
-### 3. Fix Issues in Code
-
-For each issue, provide specific code fix:
-
-**Missing Alt Text:**
-```html
-<!-- Before -->
-<img src="logo.png">
-
-<!-- After -->
-<img src="logo.png" alt="Company Logo">
-```
-
-**Poor Color Contrast:**
-```css
-/* Before - Contrast 2.8:1 (fails) */
-color: #999;
-background: #fff;
-
-/* After - Contrast 4.6:1 (passes) */
-color: #666;
-background: #fff;
-```
-
-**Missing Form Label:**
-```html
-<!-- Before -->
-<input type="email" placeholder="Email">
-
-<!-- After -->
-<label for="email">Email Address</label>
-<input type="email" id="email" aria-required="true">
-```
-
-**Keyboard Navigation:**
-```html
-<!-- Before -->
-<div onclick="handleClick()">Click me</div>
-
-<!-- After -->
-<button onclick="handleClick()">Click me</button>
-```
-
-### 4. Verify Fixes
-
-After user implements fixes:
-```
-"Re-scan [URL] to verify accessibility issues are fixed"
-```
-
-### 5. Report Results
-
-Confirm what was fixed and any remaining issues.
-
-## Quick Reference
-
-**Common WCAG Requirements:**
-- Images need alt text
-- Color contrast ≥ 4.5:1 for text
-- All interactive elements keyboard accessible
-- Forms have labels
-- Proper heading hierarchy (h1 → h2 → h3)
-
-## Example
-
-**User**: "Fix accessibility issues on my login page at localhost:3000/login"
-
-**Workflow**:
-1. Scan: `"Run accessibility scan on localhost:3000/login"`
-2. Identify: "Found 3 critical issues: missing form labels, poor contrast, no keyboard nav"
-3. Fix: Provide code fixes for each issue
-4. Verify: `"Re-scan localhost:3000/login"`
-5. Confirm: "All critical issues resolved ✓"
diff --git a/skills/run-mobile-tests-on-browserstack/SKILL.md b/skills/run-mobile-tests-on-browserstack/SKILL.md
new file mode 100644
index 0000000..fedbfca
--- /dev/null
+++ b/skills/run-mobile-tests-on-browserstack/SKILL.md
@@ -0,0 +1,112 @@
+---
+name: run-mobile-tests-on-browserstack
+description: Run automated mobile app tests on BrowserStack App Automate using runAppTestsOnBrowserStack and takeAppScreenshot MCP tools. Use for Appium/XCUITest/Espresso testing.
+---
+
+# Run Mobile Tests on BrowserStack
+
+## When to Use
+
+- Testing mobile apps on real iOS and Android devices
+- Running Appium, XCUITest, or Espresso tests
+- CI/CD integration for mobile app testing
+- Device-specific bug reproduction
+
+## MCP Tools Used
+
+- `takeAppScreenshot` - Launch app and capture verification screenshot
+- `runAppTestsOnBrowserStack` - Run automated mobile tests on real devices
+- `getFailureLogs` - Retrieve error logs for failed app tests
+
+## Steps
+
+### 1. Verify App Launch (Optional)
+
+Use `takeAppScreenshot` to verify your app launches correctly:
+
+```
+"Take a screenshot of my app on iPhone 15 Pro with iOS 17. App path: /Users/me/app.ipa"
+"Verify my app launches on Google Pixel 6 with Android 12. App file: /Users/me/app-debug.apk"
+```
+
+**What the tool does:**
+- Launches your app on specified device
+- Captures a quick screenshot for verification
+- Returns screenshot to confirm app loads correctly
+
+### 2. Run Automated Tests
+
+Use `runAppTestsOnBrowserStack` to execute your test suite:
+
+```
+"Run Espresso tests from /tests/checkout.zip on Galaxy S21 and Pixel 6 with Android 12. App path: /apps/beta-release.apk"
+"Run XCUITest tests on iPhone 14 and iPhone 15 Pro with iOS 17. Test suite: /tests/login-tests.zip"
+"Run my Appium tests on Samsung Galaxy S23 (Android 14). App: /apps/myapp.apk, Tests: /tests/suite.zip"
+```
+
+**What the tool does:**
+- Uploads your app to BrowserStack (gets bs:// URL)
+- Uploads your test suite
+- Runs tests on specified real devices
+- Returns session IDs and test results
+
+**Supported Test Frameworks:**
+- Espresso (Android native)
+- XCUITest (iOS native)
+- Appium (cross-platform)
+
+### 3. Debug Failures
+
+Retrieve error logs using `getFailureLogs`:
+
+```
+"Get error logs from App Automate session ID abc123, Build ID xyz789"
+"Show me failure logs from my last app test run"
+```
+
+## Common Scenarios
+
+**Quick App Verification:**
+```
+"Take a screenshot of my app on iPhone 15 Pro to verify it launches"
+"Check if my app opens correctly on Galaxy S24 with Android 14"
+```
+
+**Run Test Suites:**
+```
+"Run Espresso tests on Google Pixel 7 and Samsung Galaxy S23"
+"Run XCUITest suite on iPhone 14 Pro and iPhone 15"
+"Run Appium tests across 5 Android devices"
+```
+
+**Device Matrix Testing:**
+```
+"Run my tests on iPhone 14, 15, and 15 Pro with iOS 17"
+"Test on Galaxy S21, S22, S23 with Android 13 and 14"
+```
+
+**Debug Test Failures:**
+```
+"Get error logs and screenshots from my failed app test session"
+"Show me what went wrong in session abc123"
+```
+
+## Example Workflow
+
+**User**: "I need to test my checkout flow on iPhone 15 and Galaxy S24"
+
+**Steps**:
+1. Verify app: `"Take screenshot of my app on iPhone 15 Pro. App: /apps/myapp.ipa"`
+2. Run tests: `"Run XCUITest tests on iPhone 15 Pro and Appium tests on Galaxy S24. App: /apps/myapp.ipa, Tests: /tests/checkout.zip"`
+3. Tool executes: `runAppTestsOnBrowserStack` uploads app, runs tests on both devices
+4. View results and session IDs
+5. If failures: `"Get error logs from the failed sessions"`
+
+## Quick Tips
+
+- Provide full file paths for app (.ipa or .apk) and test suite (.zip)
+- Specify device name, OS, and version for precise targeting
+- Use `takeAppScreenshot` first to verify app launches
+- Session IDs and Build IDs are returned for debugging
+- Supports Espresso, XCUITest, and Appium test frameworks
+- Tests run on real physical devices, not emulators
diff --git a/skills/run-web-tests-on-browserstack/SKILL.md b/skills/run-web-tests-on-browserstack/SKILL.md
new file mode 100644
index 0000000..21f40c1
--- /dev/null
+++ b/skills/run-web-tests-on-browserstack/SKILL.md
@@ -0,0 +1,99 @@
+---
+name: run-web-tests-on-browserstack
+description: Run automated web tests on BrowserStack Automate using setupBrowserStackAutomateTests MCP tool. Use for cross-browser testing with Selenium/Playwright/Cypress.
+---
+
+# Run Web Tests on BrowserStack
+
+## When to Use
+
+- Running web tests across multiple browsers
+- Setting up CI/CD pipeline with BrowserStack
+- Testing on real desktop browsers (Chrome, Firefox, Safari, Edge)
+- Debugging cross-browser compatibility issues
+
+## MCP Tools Used
+
+- `setupBrowserStackAutomateTests` - Integrate BrowserStack SDK and run tests
+- `fetchAutomationScreenshots` - Get screenshots from test sessions
+- `getFailureLogs` - Retrieve error logs for failed tests
+
+## Steps
+
+### 1. Setup and Run Tests
+
+Use the `setupBrowserStackAutomateTests` MCP tool:
+
+```
+"Run my Selenium tests on Chrome and Firefox using BrowserStack"
+"Setup my Playwright tests for BrowserStack and run on Safari 17"
+"Run my Cypress tests on Chrome 120 (Windows 11) and Edge (Windows 11)"
+```
+
+**What the tool does:**
+- Integrates BrowserStack SDK into your test framework
+- Configures browser capabilities automatically
+- Runs tests on BrowserStack infrastructure
+- Optionally enables Percy for visual testing
+
+**Example:**
+```
+"Run my Selenium-JUnit5 tests written in Java on Chrome and Firefox. Enable Percy for visual testing."
+```
+
+### 2. Get Screenshots
+
+Fetch screenshots from test sessions using `fetchAutomationScreenshots`:
+
+```
+"Get screenshots from Automate session ID abc123xyz"
+"Show me screenshots from my last test run"
+```
+
+### 3. Debug Failures
+
+Retrieve error logs using `getFailureLogs`:
+
+```
+"Get error logs from session ID 21a864032a7459f1e7634222249b316759d6827f"
+"Show me failure logs from my last Automate test"
+```
+
+## Common Scenarios
+
+**Setup Different Test Frameworks:**
+```
+"Setup my Playwright tests for BrowserStack Automate"
+"Run my Selenium WebDriver tests on Chrome, Firefox, and Safari"
+"Configure my Cypress tests to run on BrowserStack"
+```
+
+**Enable Visual Testing:**
+```
+"Run my tests on BrowserStack with Percy enabled for visual regression testing"
+```
+
+**Debug Failed Tests:**
+```
+"Get screenshots and error logs from session abc123"
+"Show me what went wrong in my last Automate test run"
+```
+
+## Example Workflow
+
+**User**: "I need to run my Playwright tests on Chrome, Firefox, and Safari"
+
+**Steps**:
+1. Setup: `"Setup my Playwright tests for BrowserStack and run on Chrome 120, Firefox 122, Safari 17"`
+2. Tool executes: `setupBrowserStackAutomateTests` configures and runs tests
+3. View results from BrowserStack dashboard or test output
+4. If failures: `"Get error logs from the failed session"`
+5. Debug: `"Show me screenshots from session XYZ"`
+
+## Quick Tips
+
+- Specify test framework (Selenium/Playwright/Cypress) in your prompt
+- Include browser names and versions for precise targeting
+- Use Percy integration for visual regression testing
+- Session IDs are returned after test runs for debugging
+- Screenshots and logs help diagnose cross-browser issues
diff --git a/skills/scan-and-fix-accessibility/SKILL.md b/skills/scan-and-fix-accessibility/SKILL.md
new file mode 100644
index 0000000..43cae4d
--- /dev/null
+++ b/skills/scan-and-fix-accessibility/SKILL.md
@@ -0,0 +1,149 @@
+---
+name: scan-and-fix-accessibility
+description: Scan webpage for accessibility issues using startAccessibilityScan MCP tool, identify WCAG violations, and generate code fixes. Use for WCAG compliance and accessibility bug fixes.
+---
+
+# Scan and Fix Accessibility
+
+## When to Use
+
+- User wants to fix accessibility issues on a page
+- Need to ensure WCAG compliance
+- Fixing specific accessibility bugs
+
+## MCP Tools Used
+
+- `startAccessibilityScan` (Tool #17) - Scan webpage for accessibility issues
+- `accessibilityExpert` (Tool #16) - Get WCAG guidelines and best practices
+
+## Steps
+
+### 1. Run Accessibility Scan
+
+Use `startAccessibilityScan` to scan the webpage:
+
+```
+"Run accessibility scan for 'www.example.com'"
+"Scan accessibility issues on localhost:3000"
+"Run accessibility scan for 'https://mysite.com/checkout'"
+```
+
+**What the tool does:**
+- Performs comprehensive WCAG 2.0/2.1/2.2 accessibility scan
+- Returns result link with detailed violation report
+- Categorizes issues by severity (Critical/High/Medium/Low)
+
+### 2. Analyze Results
+
+Review the scan results and categorize issues:
+- **Critical**: Blocks users (missing form labels, keyboard traps)
+- **High**: Major barriers (poor color contrast, missing alt text)
+- **Medium**: Best practices (heading hierarchy)
+
+### 3. Get WCAG Guidelines (Optional)
+
+Use `accessibilityExpert` for specific guidance:
+
+```
+"What WCAG guidelines apply to form field error messages on mobile web?"
+"How do I fix color contrast issues for WCAG 2.1 AA compliance?"
+```
+
+**What the tool does:**
+- Provides WCAG 2.0/2.1/2.2 compliance guidance
+- Answers questions about accessibility best practices
+- Covers mobile and web usability standards
+
+### 4. Fix Issues in Code
+
+For each issue, provide specific code fix:
+
+**Missing Alt Text:**
+```html
+<!-- Before -->
+<img src="logo.png">
+
+<!-- After -->
+<img src="logo.png" alt="Company Logo">
+```
+
+**Poor Color Contrast:**
+```css
+/* Before - Contrast 2.8:1 (fails) */
+color: #999;
+background: #fff;
+
+/* After - Contrast 4.6:1 (passes) */
+color: #666;
+background: #fff;
+```
+
+**Missing Form Label:**
+```html
+<!-- Before -->
+<input type="email" placeholder="Email">
+
+<!-- After -->
+<label for="email">Email Address</label>
+<input type="email" id="email" aria-required="true">
+```
+
+**Keyboard Navigation:**
+```html
+<!-- Before -->
+<div onclick="handleClick()">Click me</div>
+
+<!-- After -->
+<button onclick="handleClick()">Click me</button>
+```
+
+### 4. Verify Fixes
+
+After user implements fixes, re-scan using `startAccessibilityScan`:
+
+```
+"Re-scan www.example.com to verify accessibility issues are fixed"
+"Run accessibility scan again on localhost:3000 after fixes"
+```
+
+### 5. Report Results
+
+Confirm what was fixed and any remaining issues.
+
+## Common Scenarios
+
+**Scan and Fix Workflow:**
+```
+"Scan accessibility issues on localhost:3000"
+→ Returns scan results with violations
+→ Provide code fixes for each issue
+→ User implements fixes
+"Re-scan localhost:3000 to verify fixes"
+```
+
+**Get WCAG Guidance:**
+```
+"What WCAG guidelines apply to form field error messages?"
+"How do I ensure my color palette is WCAG 2.1 AA compliant?"
+```
+
+## Quick Reference
+
+**Common WCAG Requirements:**
+- Images need alt text (WCAG 1.1.1)
+- Color contrast ≥ 4.5:1 for text (WCAG 1.4.3)
+- All interactive elements keyboard accessible (WCAG 2.1.1)
+- Forms have labels (WCAG 3.3.2)
+- Proper heading hierarchy (WCAG 1.3.1)
+
+## Example
+
+**User**: "Fix accessibility issues on my login page at localhost:3000/login"
+
+**Workflow**:
+1. Scan: Use `startAccessibilityScan` for "localhost:3000/login"
+2. Results: Scan returns violations - 3 critical issues found
+3. Identify: "Missing form labels, poor contrast, no keyboard nav"
+4. Fix: Provide code fixes for each issue
+5. Verify: Use `startAccessibilityScan` again to re-scan
+6. Confirm: "All critical issues resolved ✓"
diff --git a/skills/visual-issue-fix/SKILL.md b/skills/visual-issue-fix/SKILL.md
deleted file mode 100644
index 64989df..0000000
--- a/skills/visual-issue-fix/SKILL.md
+++ /dev/null
@@ -1,130 +0,0 @@
----
-name: visual-issue-fix
-description: Capture screenshots of a page across browsers, identify visual issues, fix them in code, and validate. Use for cross-browser layout bugs or visual regressions.
----
-
-# Visual Issue Fix Workflow
-
-## When to Use
-
-- Layout broken in specific browser
-- Visual bugs reported by users
-- Cross-browser design inconsistencies
-- Responsive design issues
-
-## Steps
-
-### 1. Capture Screenshots
-
-Open page on target browsers:
-```
-"Open [URL] on Chrome"
-"Open [URL] on Safari"
-"Open [URL] on Firefox"
-```
-
-Take screenshots:
-```
-"Take screenshot of the current page"
-```
-
-### 2. Identify Issues
-
-Compare screenshots and identify:
-- Layout differences (alignment, spacing)
-- Missing elements
-- Broken styling (fonts, colors, borders)
-- Responsive breakpoint problems
-
-### 3. Fix in Code
-
-Provide specific code fixes:
-
-**Flexbox Issue:**
-```css
-/* Before - Safari issue */
-.container {
-  display: flex;
-  gap: 20px;
-}
-
-/* After - Works everywhere */
-.container {
-  display: flex;
-}
-.container > * {
-  margin-right: 20px;
-}
-```
-
-**CSS Grid Issue:**
-```css
-/* Before - IE11 doesn't support */
-.grid {
-  display: grid;
-  grid-template-columns: repeat(3, 1fr);
-}
-
-/* After - Fallback for old browsers */
-.grid {
-  display: flex;
-  flex-wrap: wrap;
-}
-@supports (display: grid) {
-  .grid {
-    display: grid;
-    grid-template-columns: repeat(3, 1fr);
-  }
-}
-```
-
-**Responsive Issue:**
-```css
-/* Before - Fixed width breaks on mobile */
-.sidebar {
-  width: 300px;
-}
-
-/* After - Responsive */
-.sidebar {
-  width: 100%;
-}
-@media (min-width: 768px) {
-  .sidebar {
-    width: 300px;
-  }
-}
-```
-
-### 4. Validate Fix
-
-After user implements fix:
-```
-"Re-open [URL] on [browser] to verify the fix"
-"Take screenshot to confirm layout is correct"
-```
-
-### 5. Test Other Browsers
-
-Ensure fix didn't break other browsers:
-```
-"Test on Chrome, Firefox, Safari, and Edge"
-```
-
-## Quick Tips
-
-- Test on real devices, not just dev tools
-- Check both desktop and mobile
-- Verify at different screen sizes
-- Look for vendor prefix needs (-webkit-, -moz-)
-
-## Example
-
-**User**: "My navigation is broken in Safari"
-
-**Workflow**:
-1. Capture: `"Open localhost:3000 on Safari"` + `"Take screenshot"`
-2. Identify: "Navigation items overlapping - flexbox gap issue"
-3. Fix: Provide CSS fix replacing gap with margins
-4. Validate: `"Re-open localhost:3000 on Safari"` - confirm fixed
-5. Cross-check: Test on Chrome, Firefox to ensure no regression